This is the first report of a comparative proteomic analysis of vitreous humour from type 2 diabetic patients with PDR with that from normal human eyes donated for corneal transplant. From the vitreous samples, we detected more than a thousand spots at a time on a single gel using DIGE; of these spots, 57 showed highly significant changes in the expression level compared with the control group and were successfully analysed as corresponding to peptide fragments of 29 proteins, including 8 proteins that increased and 21 proteins that decreased in PDR. Excluding the serum proteins from minor vitreous haemorrhage, 19 proteins were identified as differentially produced in the vitreous fluid of the PDR patients compared with the vitreous fluid of the normal subjects; 6 of these proteins have never been reported to be differentially expressed in the PDR vitreous: N(G),N(G)-dimethylarginine dimethylaminohydrolase 1 (DDAH 1), tubulin alpha-1B chain, gamma-enolase, cytosolic acyl coenzyme A thioester hydrolase (ACOT1), malate dehydrogenase (MDH) and phosphatidylethanolamine-binding protein 1 (PEBP1). The observed differences in PEDF and clusterin levels by DIGE were further validated by western blot analysis of the samples from each patient, which confirmed the observed differences with good quantitative agreement. The results could not be extrapolated to the other candidates because they have not been verified by Western blot analysis.
Some of the identified proteins appeared at multiple positions on the gels in this and other studies [6, 16], which is consistent with the presence of different post-translationally modified forms. Post-translational modifications can change the MW and the pI of proteins, and the various forms of these proteins can migrate to different spots. The quantification of the identified proteins was based on the intensities from multiple spots.
PEDF is produced by the retinal pigment epithelium and serves as a major inhibitor of intraocular angiogenesis. There is growing evidence to suggest that PEDF has a modulatory role in angiogenesis . PEDF alterations in patients with PDR compared with nondiabetic patients are controversial. Some previous studies pointed to reductions in the levels of vitreous PEDF in patients suffering from PDR [6, 18–20]. Conversely, elevated levels of PEDF were detected in some studies [13, 21]. We have detected reduced levels of PEDF in the vitreous fluid of diabetic patients with PDR by both proteomic analysis and Western blotting. PEDF might be candidate target protein for diabetic retinopathy treatment.
Clusterin is a secreted glycoprotein that has been implicated in a variety of physiological processes, including cell-cell interaction, lipid transport, tissue remodelling, chaperone activity, and apoptosis [22, 23]. In recent years, clusterin has been considered a potential diagnostic and prognostic biomarker for several human cancers [24–27]. An and colleagues have demonstrated that clusterin is produced and secreted by retinal pigment epithelial (RPE) cells . Previous studies suggest that during diabetes-induced retinal damage, cytoplasmic clusterin is likely to be associated with protection from cell death, while nuclear clusterin might promote cell death . It is known that clusterin interacts with TGF-β type II receptor (Tβ R II) , and TGF-β plays multifunctional roles in regulating the cell cycle, apoptosis, differentiation, and extracellular matrix remodelling . Clusterin is also an important mediator of cell signalling; it can interfere with NF-κB, PI3 kinase, or MAP kinase signalling [32, 33], which are associated with cell apoptosis and cell proliferation. In a mouse model of DR, clusterin reduced the leakage from vessels in the diabetic retina, which was accompanied by the restoration of the expression of tight junction proteins [34, 35]. These observations suggested that clusterin may play an important role in the prevention of diabetes-induced BRB breakdown. Our data show the downregulation of clusterin in the PDR vitreous, which was confirmed by Western blotting; however, the function of clusterin in the vitreous is not yet clear. PDR is characterised by neovascularisation and enhanced vascular permeability. The proteins involved in the regulation of cell proliferation, apoptosis and BRB breakdown may play important roles in PDR pathogenesis. Therefore, clusterin may contribute to the pathogenesis of PDR, and further studies investigating the precise role of clusterin in diabetic retinopathy are needed.
Carbonic anhydrase (CA) has been identified by proteomic analysis in the vitreous humour from PDR patients; CA production was increased in the vitreous of diabetic retinopathy patients compared with the controls [7, 8]. In this study, we have confirmed this finding. The presence of CA was initially thought to be due to retinal haemorrhage and erythrolysis, but subsequent investigation demonstrated that CA is also actively involved in the progression of diabetic retinopathy; CA-1 expression leads to the activation of the contact system, the intrinsic pathway of coagulation, and promotes retinal vessel leakage and intraretinal edema via increased kallikrein activity . The mechanisms involved in the increasing production of CA and the value of CA to potential therapeutic targets require further investigation.
In this study, we identified six proteins that have not previously been reported or described in PDR. It is noteworthy that four of them (DDAH, gamma-enolase, cytosolic acyl coenzyme A thioester hydrolase and malate dehydrogenase) are cellular enzymes, and their levels are all decreased in the vitreous of diabetic patients with PDR, but we do not know whether these changes represent primary causes or consequences of PDR. DDAH is an extremely oxidant-sensitive enzyme . Decreased DDAH expression/activity is evident in disease states associated with endothelial dysfunction and is believed to be the mechanism responsible for the increase in methylarginine levels and the subsequent asymmetric dimethylarginine (ADMA)-mediated endothelial nitric oxide synthase impairment [37, 38]. It should be noted that vascular endothelial cells are major targets of hyperglycaemic damage; endothelial dysfunction and reduced levels of endothelial progenitor cells can cause microvascular complications in diabetes mellitus [39, 40]. Additionally, oxidative stress appears to play an important role in endothelial dysfunction in diabetes. Therefore, it is thought that the DDAH/ADMA pathway can potentially modulate NO production and endothelial function in PDR. In addition, gamma-enolase has been found in different types of human cancer and is used as a marker for tumoural or cellular damage [41, 42]. Gamma-enolase is also used as a marker for neural damage and is a reliable marker for cellular stress during rhegmatogenous retinal detachment (RRD) [43, 44]. The changed expression levels of cytosolic acyl coenzyme A thioester hydrolase and malate dehydrogenase in the PDR vitreous could reflect the alterations in glucose and lipid metabolism [45, 46]. Moreover, acyl-CoA thioesterases are highly regulated by peroxisome proliferator-activated receptors (PPARs) , and PPARγ agonists have shown promise as targets in animal models of proliferative retinopathies . Further studies investigating the role of these enzymes in diabetic retinopathy are needed.
PEBP is a protease inhibitor, and it has been demonstrated to bind to Raf-1 and mitogen-activated protein kinase (MAPK), components of the extracellular signal-regulated protein kinase (ERK) pathway . Aberrant signalling through the ERK pathway could promote cell immortalisation via such mechanisms as telomerase induction, growth factor-independent proliferation, and angiogenesis by the upregulation of proangiogenic factors . Therefore, the lower levels of PEBP detected in PDR patients are perhaps related to neovascularisation and cell cycle progression.
Several types of crystallins, including beta-crystallin S, beta-crystallin B2, alpha-crystallin B chain, beta-crystallin A4, beta-crystallin A3, and gamma-crystallin C, were found in the vitreous humour of both PDR patients and controls. All types of crystallins found in this study were significantly lower in the vitreous humour from PDR patients compared with that from the control subjects, and a previous study has reported that beta-crystallin B2 was identified by MALDI-TOF in normal vitreous . We do not know why the crystallin levels are decreased in the vitreous of diabetic patients with PDR; the biological functions of crystallins are not completely understood. However, it is noteworthy that αA-crystallin and advanced glycation end product (AGE) were highly expressed in human diabetic retinas, and αA-crystallin expression was up-regulated in murine posterior eyecups after recombinant AGE protein was injected into the vitreous of adult murine eyes, αA-crystallin responded to AGE accumulation, which may contribute to the protection of photoreceptors against AGE-related retinal tissue injury . Thus, the mechanisms involved in the intraocular production of crystallins and the role of crystallins in the pathogenesis of PDR require further investigation.
There are two main limitations to this study. First, the gel electrophoresis technique has a number of significant drawbacks. These include its inability to detect low-abundance proteins in the presence of high-abundance proteins or to separate proteins that are too basic, too acidic, too large, or too small [52, 53]. Thus, the detection shortcoming may be responsible for our failure to detect vascular endothelial growth factor (VEGF), a key mediator of retinal neovascularisation and vascular permeability in the pathogenesis of diabetic retinopathy . We will apply other proteomic technologies in an effort to pursue this issue. Second, vitreous haemorrhage that occurs in PDR can produce a massive influx of serum proteins. Although we excluded any samples displaying gross vitreous haemorrhage, our results also included some serum proteins resulting from minor vitreal haemorrhage or leakage of serum into the vitreous.